I n the past two decades, the total number of global migrants has increased by nearly 50 percent to 258 million. In response to this increase, many countries hosting migrants have focused on increasing border security and regulating migration. In the United States, home to around 50 million migrants, increased border security has been a key focus of the Trump administration. Between 2004 and 2017, the U.S. Customs and Border Protection’s enacted budget increased from $6 billion to $14.3 billion, a 138 percent increase. Since the passage of the 2006 Secure Fence Act, a key element of efforts to increase border enforcement has been the construction of a wall along the U.S.-Mexico border.

While the costs of such infrastructure investments are known (and are economically significant), the benefits of border wall expansion are difficult to quantify. Estimates suggest that completing a physical barrier along the entire U.S.-Mexico border would cost around $60 billion. Proposed benefits by proponents of border wall expansion can be classified into labor market effects (wage and employment effects on native populations from reduced migration), public expenditure effects (changes in public program spending from changes in migrant flows), and effects on crime and safety. Recent estimates find that expansion of border walls along the U.S.-Mexico border between 2007 and 2010 harmed both Mexican workers and U.S. workers in high‐​skilled occupations. U.S. low‐​skill workers had welfare gains equivalent to a $0.28 increase in per capita income, far below the per capita cost of border wall construction.

Beyond the labor market effects of border walls, recent public discourse has focused on the crime and safety effects of these barriers, especially after Trump during his State of the Union address in 2019 made the unsubstantiated claim that the border wall had greatly reduced crime in the city of El Paso, Texas. While there is no shortage of opinions on crime in border regions, there is little rigorous empirical evidence on the potential impact of border infrastructure on local crime.

Our research addresses this gap in the literature and in the political discourse. We examine the unprecedented expansion of border infrastructure built along the U.S.-Mexico border following the passage of the 2006 Secure Fence Act and its effects on county‐​level property and violent crime rates. We conduct our analysis in two ways: one that allows us to look at differences in crime rates for each county individually and one that allows us to look for overall patterns in crime rates around border wall construction. Our analysis evaluates the causal effects of border wall construction on crime by comparing changes in crime in counties that had border wall construction between 2006 and 2012 to demographically and economically similar counties that did not experience border wall construction. Overall, we find that construction of border fencing has no appreciable effects on either property or violent crime rates. Only when we separate the sample into border wall construction and postconstruction periods do we find any crime‐​deterrent effects of border walls. We do find a short‐​term (i.e., during the construction period) effect of 0.34 per 1,000 fewer property crimes in Yuma County, Arizona, and 1.52 per 1,000 fewer violent crimes in Luna County, New Mexico. The magnitudes of these reductions are very small, especially when compared to the cost of border wall construction. For example, there are 61.5 miles of primary pedestrian border fencing in Yuma County. A 2009 Government Accountability Office report estimated that pedestrian fencing built following the 2006 Secure Fence Act cost an average of $3.9 million per mile, suggesting that it cost around $240 million to build Yuma County’s pedestrian border wall with Mexico. Can this expense be justified by the 0.34 fewer property crimes per 1,000 we estimate as a result of the border wall? Similarly, can the roughly $226 million it cost to build the 58 miles of total border barriers in Luna County be sufficient justification for the reduction of 1.52 violent crimes per 1,000? Furthermore, the apparent decline in violent crime in Luna County during the construction period is more than offset by an estimated increase in crime after construction was completed.

We combine the analyses to more explicitly test for patterns of declining crime rates concurrent with or shortly following additional barrier construction. We find no such patterns in property crime, and we find that short decreases in violent crime actually precede barrier construction rather than follow it. This suggests that any reductions in crime are likely to have been driven by the short‐​term economic stimulus provided by border wall construction. Overall, these findings indicate no systematic relationship between border wall construction and crime rates.

NOTE:
This research brief is based on Ryan Abman and Hisham Foad, “Border Walls and Crime: Evidence from the Secure Fence Act,” November 2019, https://​papers​.ssrn​.com/​s​o​l​3​/​p​a​p​e​r​s​.​c​f​m​?​a​b​s​t​r​a​c​t​_​i​d​=​3​4​82004.